Our group explores synergies between statistical science and machine learning, as applied to quantitative large-scale mass spectrometry-based investigations, to understand the functioning of living organisms. Our work covers (1) statistical experimental design, (2) detecting analyte’s signals in large and complex outputs produced by the instruments, (3) statistical selection of the relevant signals, and (4) inference of causal associations among the analytes that are informative of their biological function.
We develop methods and open-source software, that are broadly used in academia and industry. These include MSstats (relative quantification of proteins in mass spectrometric experiments) and Cardinal (analysis of mass spectrometric images). This also includes the infrastructure within the public repository MassIVE.quant for reproducible documentation and reanalyses of quantitative proteomic experiments. Our work is recognized with the 2019 Chan Zuckerberg Essential Open Source Software Award, and with the 2021 Gilbert S. Omenn Computational Proteomics Award of the US Human Proteome Organization.